Skip to content

535 Final forecast · Nov 3
Dispatches

More

← All dispatches

Poll reaction FL-14

Emerson poll shows a closer FL-14 race than our model, which barely moves

A small Emerson College survey has Kathy Castor up 48 to 43 over Mike Beltran, well below our model's expected margin, and the forecast shifts by a tenth of a point.

Written by the newsroom · anthropic/claude-opus-5.5 · October 22, 2026 at 9:03 AM EDT

Correction: the House forecasts used in this article predate our redistricting correction. They used 2024 House results on earlier district lines. The current forecasts use presidential results recalculated on the 2026 lines for affected districts. Read the methodology.

Correction: the forecasts in this article predate a polling-data repair. Some imported polls omitted candidates who shared a party, and ranked-choice rounds were not consistently selected. The current forecasts use corrected results and recalculated pollster adjustments. See current forecasts.

A new Emerson College poll of Florida's 14th District, sponsored by Poll Vault, finds Democratic Rep. Kathy Castor at 48 percent and Republican Mike Beltran at 43 percent. The survey of 315 likely voters was fielded October 12-14. The payload does not report its margin of error, and the sample is small.

The poll is far closer than our model. Our forecast puts Castor's average margin at D+26.6, with a 5th-to-95th percentile range of D+13.8 to D+39.3. The Emerson result, a 5-point gap, falls below the low end of that range.

The new numbers did not change the picture. Our model gives Castor a 100 in 100 chance of winning the seat, the same as in the run from 8:32 AM ET today. Her average margin moved from D+26.5 to D+26.6, and both ends of the range shifted by a tenth of a point. One small survey that disagrees with the rest of the model's inputs carries limited weight on its own.

Nationally, our model gives Democrats a 99% chance of winning a House majority, which takes 218 seats, with an average of 246.6 Democratic seats. That figure is marginally firmer than it should be. Our simulation correlates error nationally, by census division, by state and by district, but it does not link demographically similar districts in different places, because we lack the data to do so.

Polls this dispatch was written from

  • Emerson College · Oct 14, 2026 · n=315 · FL-14

Every constant behind these numbers is on the methodology page.